The Agent Skills Directory

[SAFE]: The skill consists exclusively of natural language instructions for the AI agent to conduct a structured interview. No technical security risks were identified.
The instructions explicitly forbid writing files (docs/monorail/) or implementing code, enforcing a stateless execution model.
No external dependencies, remote scripts, or network requests are defined in the skill configuration or metadata.
The disable-model-invocation: true flag in the YAML frontmatter and the allow_implicit_invocation: false policy in the agent configuration further restrict the agent's operational scope, reducing the attack surface.

rail-grill